Global, regional, and national burden of knee osteoarthritis: findings from the Global Burden of Disease study 2021 and projections to 2045

Abstract

Background

Osteoarthritis is a major contributor to the global disease burden, with the knee being the most commonly affected site. Comprehensive research on knee osteoarthritis (KOA) at a global scale carries significant implications for public health.

Methods

The estimates and 95% uncertainty intervals (UIs) for prevalence, incidence and disability-adjusted life years (DALYs) of KOA were extracted from Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2021. We described KOA epidemiology at global, regional, and national levels, analyzed 1990–2021 trends in KOA burden from overall, local, and multi-dimension scopes, decomposed KOA burden according to population size, age structure, and epidemiologic changes, quantified cross-country inequalities in KOA burden using standard health equity methods recommended by World Health Organization, and predicted changes of KOA burden to 2045.

Results

GBD 2021 estimated 374,738,744 (95% UI: 321,858,982–428,353,220) prevalent cases, 30,845,891 (95% UI: 26,534,151–35,188,905) incident cases and 12,019,070 (95% UI: 67.08–266.87) DALYs cases of KOA worldwide in 2021. An overall increase in prevalence rates, incidence rates and DALYs rates were observed from 1990 to 2021. And the prevalence number, incidence number and rates were consistently higher in females compared to males. Decomposition analysis revealed that aging, population growth, and epidemiological changes contributed 15.6%, 74.72%, and 9.69% to the global increase in age-standardized prevalence rate, respectively. A decrease in Socio-Demographic Index (SDI)-related inequalities was detected. Notably, the case number of these metrics were predicted to keeping increasing, with predicted values of 658,088,384.48 (322,110,040.98–994,066,727.98), 47,256,502.97 (23,440,017.7–71,072,988.23) and 20,517,479.78 (10,056,930.68–30,978,028.88), respectively.

Conclusions

The global burden of KOA has shown a consistent upward trend from 1990 to 2021, primarily driven by population growth and aging demographics. Countries with high SDI faced a disproportionately high burden of KOA, although inequalities related to SDI among nations have decreased over time. This study underscores significant challenges in managing KOA, including the increasing number of cases and ongoing disparities worldwide.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13018-025-06140-0.

Background

As global aging progresses, osteoarthritis (OA) has emerged as one of the most prevalent joint diseases, with increasing incidence rates over the years [1]. Knee osteoarthritis (KOA) is particularly notable as the most commonly affected site [2]. The onset and progression of KOA are influenced by various factors, including acute and chronic joint injuries, age, obesity, and metabolic bone disease [3]. The pathogenesis involves cartilage degeneration, bone remodeling, osteophyte formation, and chronic inflammation, leading to symptoms such as knee pain, limited mobility, dysfunction, joint deformity, and even disability [4, 5]. In clinical practice, multiple treatment options exist for KOA pain, including basic care, pharmacological interventions, and surgical procedures [6]. Notably, knee arthroplasty stands out as the only effective option for advanced stages, incurring significant health costs [7].

OA represents a major source of healthcare expenditure, with an estimated global prevalence of approximately 364.6 million cases of KOA in 2019, accounting for 4.9% of all global health issues [8]. KOA contributes to considerable health and economic burdens, encompassing not only direct and indirect costs but also intangible factors such as fatigue and decreased social participation [9]. Given the trends in an aging population and the obesity epidemic, there is an urgent need to address the disease burden of KOA [10].

Recognizing KOA as a chronic, progressive disorder with irreversible structural changes emphasizes the importance of early proactive management. However, barriers such as the lack of validated early-stage diagnostic criteria, insufficient evidence for the efficacy of non-surgical treatments, and limited public health awareness—especially in underdeveloped regions—hinder effective management and prevention [11]. The Global Burden of Disease (GBD) study offers a systematic approach to estimating the burden of KOA at global and regional levels [9]. Previous studies have examined the worldwide epidemiology of OA using GBD data, with the 2021 study providing updated insights into the global, regional, and national burden of OA from 1990 to 2020 [12]. Although findings indicate a continuing rise in age-standardized years lived with disability due to OA, a comprehensive analysis of the global burden of KOA utilizing the latest GBD 2021 data remains absent.

To enhance understanding of KOA epidemiology globally, this study leverages data from the GBD 2021 to provide an updated, comprehensive evaluation of the burden, trends, and inequalities associated with KOA. This aims to identify the most affected populations and inform targeted prevention and treatment strategies.

Materials and methods

Data acquisition

The GBD 2021 study provides a thorough evaluation of health loss associated with 369 diseases, injuries, and impairments, as well as 88 risk factors, across 204 countries and territories. It leverages the most recent epidemiological data and refined standardized methods to ensure robust findings [13]. The GBD database utilizes advanced techniques to address missing data and adjust for confounding variables. Comprehensive details about the study design and methodologies are extensively documented in prior GBD literature [14]. For this study, estimates and their 95% uncertainty intervals (UIs) for the prevalence, incidence, and disability-adjusted life years (DALYs) associated with KOA were derived from the GBD 2021 dataset. The study also utilized the sociodemographic index (SDI), which measures a country’s or region’s development level based on fertility rates, education levels, and per capita income [15]. The SDI ranges from 0 to 1, with higher values indicating greater socioeconomic development. Since the SDI is known to correlate with disease incidence and mortality rates, we classified countries and regions into five SDI categories (low, low-medium, medium, medium–high, and high) to explore the relationship between KOA burden and socioeconomic development.

Burden description

In 2021, a comprehensive assessment was conducted to quantify the national burden of KOA, including its prevalence, incidence, and DALYs. The study also explored demographic factors influencing the impact of KOA, analyzing the distribution of the disease burden across different age groups and genders.

Joinpoint regression analysis

This study employed the Joinpoint regression analysis model, a statistical technique frequently used in epidemiological research to assess temporal trends in disease prevalence or mortality. The model effectively identifies and characterizes significant change points within time-series data, here applied to KOA prevalence across global, continental, and national scopes. A log-linear model was used to perform segmented regression, and the grid search method was employed to generate all possible joinpoints. For each potential configuration, the mean squared error was calculated, and the grid point with the lowest the mean squared error was selected as the joinpoint. Subsequently, the optimal number of joinpoints in the regression model was determined using the Monte Carlo permutation test. Finally, the annual percent change (APC) and the average annual percent change (AAPC) derived from the optimal model were used to quantify the trend in rates from 1990 to 2021.

Age-period-cohort analysis

The age-period-cohort model was employed to analyze trends in KOA prevalence, incidence and DALYs. This model, commonly used to assess changes and causes of prevalence, incidence and mortality in chronic diseases, helps describe disease trends based on the effects of age, period, and cohort, and also allows for trend prediction. In the age-period-cohort model, there is a complete linear relationship, “period = age + cohort,” among the three factors, which can complicate parameter estimation. This study addressed the issue by developing estimable age-period-cohort parameters and functions without imposing arbitrary constraints on model parameters. The age-period-cohort model was implemented using R tools, with detailed methods described in previous literature [16–19].

The analysis utilized estimated prevalence, incidence and DALYs values for KOA from GBD 2021, along with population data, as input for the age-period-cohort model. Data for 16 age groups (30–34, 35–39, …, 95 plus) were structured into single units representing characteristic ages, and data for six time periods (1992–1996, 1997–2001, …, 2017–2021) were similarly structured to represent specific periods. A fitted age-period-cohort model was used to estimate the net change in KOA prevalence, incidence, and DALYs, accounting for age, period, and cohort effects. The age-period-cohort model calculated the annual percentage change in age-specific prevalence, incidence, and DALYs, using a randomly selected group as the reference period (cohort) [20]. Relative risk was employed to calculate the age-specific incidence rate ratio of each period (cohort) to the reference period (cohort). The Wald χ² test was applied to assess statistical significance, with p values < 0.05 considered statistically significant. All analyses were performed using R (4.2.1).

Decomposition analysis

Decomposition analysis is a method used to determine how various factors contribute to overall differences in values. This approach helps uncover substantial heterogeneity in demographic and epidemiological trends [21]. To quantify the relative contributions of changes in age structure, population size, and epidemiology to the overall burden of KOA from 1990 to 2021, we employed Das Gupta’s decomposition method [22]. Specifically, the number of DALYs (and similarly, prevalence and incidence) for a given year y can be expressed as: ${\mathit{DALYs}}_{ay,py,ey}={\sum }_{i=1}^n\left(a_{i,y},,\ast ,p_y,,\ast ,e_{i,y}\right)$, where a_i,y is the proportion of the population in age group i in year y, p_y is the total population in year y, and e_i,y is the age-specific DALY rate for group i in year y. In this formulation, DALYs_{ay, py, ey} represents the total number of DALYs determined by the joint effects of age structure, population size, and DALY rates [11]. Changes in age structure primarily reflect population aging, changes in population size reflect demographic growth or decline, and epidemiological changes correspond to shifts in age-standardized incidence and mortality rates [23]. The individual contribution of each factor to the total change in DALYs from 1990 to 2021 was estimated by varying one component at a time while holding the others constant—thereby isolating the effect of age structure, population growth, or epidemiological change, respectively.

Cross-country inequality analysis

Monitoring health inequalities serves as a crucial basis for evidence-based health planning, enabling the enhancement of policies, programs, and practices to address disparities in health outcomes. In this study, we utilized two standard metrics—slope index of inequality and concentration index—to evaluate the distributional inequality of KOA burden across countries [11, 24]. The slope index of inequality was determined by regressing national DALY rates for all age groups against a relative position scale linked to sociodemographic development. The concentration index was derived by calculating the area under the Lorenz curve, constructed using the cumulative proportion of DALYs and the cumulative relative distribution of the population, ranked by SDI [22].

Predictive analysis

To support the formulation of more effective public health policies and improve the allocation of healthcare resources, we conducted a prediction of the global burden of knee osteoarthritis through the year 2045. For this purpose, we applied the Bayesian age-period-cohort analysis model combined with the integrated nested Laplace approximation method, which has been shown to offer better coverage and higher predictive precision than the traditional age-period-cohort model [25, 26]. This approach allows for the approximation of marginal posterior distributions without the need for Markov Chain Monte Carlo sampling, thereby avoiding common issues such as slow convergence and poor mixing that are often associated with traditional Bayesian methods [25].

Results

Global level

In 2021, the global burden of KOA remained significant, with 374,738,744 cases (95% UI: 321,858,982–428,353,220), reflecting an alarming 234.5% increase since 1990. The age-standardized prevalence rate (ASPR) rose from 3,964.75 per 100,000 persons (95% UI: 3,411.86–4,536.4) in 1990 to 4,294.27 per 100,000 (95% UI: 3,695.04–4,910.76) in 2021, with an estimated annual percentage change (EAPC) of 0.33 (95% CI: 0.3 to 0.35) (Table 1, Fig. 1A). The global incidence of KOA reached 30,845,891 cases (95% UI: 26,534,151–35,188,905) in 2021, marking a 218.23% increase since 1990, while the age-standardized incidence rate (ASIR) increased from 330.26 per 100,000 (95% UI: 284.34–375.75) in 1990 to 353.67 per 100,000 (95% UI: 304.56–402.5) in 2021, with an EAPC of 0.28 (95% CI: 0.26 to 0.3) indicating a consistent rise in incidence (Table 1, Fig. 1B). The global DALYs for KOA in 2021 totaled 12,019,070 (95% UI: 5,858,108–23,267,858), with an age-standardized DALY rate of 137.59 per 100,000 (95% UI: 67.08–266.87) and an EAPC of 0.33 (95% CI: 0.3–0.36) (Table 1, Fig. 1C).

Table 1.

Prevalent cases, incident cases and DALYs for KOA in 2021 and estimated annual percentage changes by Global Burden of Disease

Location	Prevalence (95% uncertainty interval)					Incidence (95% uncertainty interval)
	Cases (1990)	ASR (1990)	Cases (2021)	ASR (2021)	EAPC	Cases (1990)	ASR (1990)	Cases (2021)
Global (both sex)	159798909 (137277437,182882554)	3964.75 (3411.86,4536.4)	374738744 (321858982,428353220)	4294.27 (3695.04,4910.76)	0.33 (0.3,0.35)	14134284 (12150933,16079381)	330.26 (284.34,375.75)	30845891 (26534151,35188905)
Global (female)	98498538 (84692265,112495598)	4613.23 (3972.13,5267.02)	230189285 (198051498,262858482)	5029.51 (4331.89,5738.43)	0.36 (0.33,0.4)	8393838 (7253340,9518118)	4613.23 (3972.13,5267.02)	18299871 (15784959,20836893)
Global (male)	61300372 (52267191,70269784)	3224.33 (2771.36,3685.31)	144549459 (123542116,165818687)	3483.37 (2991.52,3989.67)	0.3 (0.27,0.32)	5740445 (4906389,6564723)	3224.33 (2771.36,3685.31)	12546020 (10759562,14371725)
High SDI	47279525 (41017926,53920968)	4364 (3780.18,4972.97)	90099016 (78371034,102834145)	4656.7 (4033.58,5313.55)	0.16 (0.12,0.2)	3765186 (3281345,4286856)	363.52 (313.7,414.49)	6502877 (5683246,7463504)
High-middle SDI	37990614 (32600684,43592484)	3772.7 (3248.87,4327.23)	85016578 (72744239,97535184)	4295.79 (3685.79,4926.17)	0.53 (0.48,0.59)	3321371 (2852523,3782455)	321.69 (277.21,366.51)	6758885 (5792871,7757737)
Middle SDI	44648677 (37987635,51370949)	4102.09 (3527.55,4699.44)	123424845 (105700584,141777409)	4396.08 (3789.23,5031.25)	0.38 (0.32,0.45)	4130154 (3517288,4720592)	338.37 (291.05,386.6)	10373196 (8911389,11926381)
Low-middle SDI	21885269 (18774477,25081455)	3427.13 (2946.77,3915.09)	56770333 (48884057,64923405)	3790.26 (3258.77,4325.52)	0.35 (0.34,0.36)	2125659 (1820455,2420457)	294.4 (252.81,334.62)	5255956 (4503893,5974804)
Low SDI	7844041 (6696613,8986414)	3324.31 (2859.98,3803.29)	19146655 (16425264,21911203)	3545.62 (3048.24,4056.6)	0.22 (0.21,0.23)	778637 (667577,888772)	288.29 (248.25,328.53)	1931578 (1655106,2209900)
Andean Latin America	853460 (738034,976452)	4070.11 (3516.08,4660.84)	2758218 (2370967,3163149)	4588.33 (3950.43,5244.03)	0.42 (0.41,0.43)	80832 (69708,92097)	348.24 (300.96,396.58)	242602 (209062,279223)
Australasia	993266 (867688,1128276)	4288.08 (3746.3,4863.21)	2389034 (2066114,2736476)	4808.67 (4147.42,5488.06)	0.36 (0.34,0.39)	80753 (70372,92095)	359.26 (310.77,412.04)	177000 (153385,205240)
Caribbean	1066131 (923195,1224744)	4080.07 (3540.14,4677.22)	2403792 (2070228,2740275)	4454.36 (3840.02,5084.92)	0.32 (0.3,0.33)	94617 (81658,107550)	348.79 (300.57,396.78)	199696 (172770,230459)
Central Asia	1192276 (1022977,1370278)	2533.07 (2186.47,2911.04)	2301085 (1969099,2659417)	2700.22 (2336.33,3105.34)	0.21 (0.2,0.23)	111442 (96037,128400)	225.44 (194.93,257.32)	222031 (188796,255746)
Central Europe	4512504 (3866745,5163412)	3005.2 (2586.99,3442.37)	6774519 (5844666,7842619)	3230.75 (2788.51,3718.9)	0.25 (0.24,0.25)	392665 (336818,453197)	263.75 (226.58,302.68)	529220 (459378,611643)
Central Latin America	3520770 (3020725,4018304)	4114.19 (3551.89,4687.11)	11498876 (9894419,13149726)	4492.79 (3881.3,5124.51)	0.28 (0.28,0.29)	340356 (292287,386909)	354.35 (305.58,404.67)	1014222 (871109,1156490)
Central Sub-Saharan Africa	776748 (661806,895686)	3288.2 (2816.19,3766.21)	2053774 (1756352,2357159)	3433.3 (2947.25,3926.7)	0.11 (0.09,0.13)	79094 (67356,90793)	286.39 (245.77,329.26)	217592 (185701,248119)
East Asia	42602441 (35969751,49229166)	4662.94 (3992.86,5359.08)	113375622 (96002567,130995916)	5016.78 (4267.39,5758.88)	0.49 (0.39,0.6)	3787399 (3238108,4358778)	377.47 (324.32,433.77)	8800266 (7530686,10174644)
Eastern Europe	8853765 (7574166,10111455)	3154.42 (2713.05,3603.85)	11762775 (10059205,13517483)	3409.25 (2919.69,3923.46)	0.28 (0.27,0.3)	769391 (659450,894233)	278.19 (239.66,320.43)	958135 (823739,1109492)
Eastern Sub-Saharan Africa	2491439 (2121796,2866357)	3216.44 (2769.51,3707.07)	6218769 (5299896,7142986)	3446.18 (2959.9,3965.07)	0.24 (0.23,0.26)	252480 (215755,289261)	281.41 (242.27,322.52)	652313 (555382,747784)
High-income Asia Pacific	11135071 (9603769,12692418)	5416.8 (4683,6163.06)	22540484 (19570221,25550915)	5573.73 (4841.29,6334.33)	0.15 (0.11,0.19)	928187 (799581,1062050)	441.54 (382.22,504.23)	1482839 (1312650,1690590)
High-income North America	15143603 (13138164,17316618)	4500.08 (3894.98,5121.91)	29008914 (25038463,33373205)	4709.03 (4074.67,5394.69)	-0.11 (-0.31,0.08)	1160725 (1017358,1319031)	367.48 (317.93,419.67)	2103504 (1830759,2428582)
North Africa and Middle East	5916998 (5065219,6807128)	3378.07 (2909.97,3880.84)	18591752 (15988433,21426733)	3810.43 (3275.1,4375)	0.39 (0.38,0.4)	578152 (495303,662749)	292.26 (251.18,334.1)	1811718 (1543012,2081311)
Oceania	119873 (101127,139191)	3730.88 (3201.2,4301.16)	338675 (285507,391076)	4025.21 (3433.77,4602.96)	0.2 (0.17,0.23)	11950 (10219,13792)	312.73 (269.5,359.55)	33635 (28494,39060)
South Asia	21001741 (17861857,24088981)	3441.76 (2953.66,3924.8)	58791056 (50326960,67052871)	3818.04 (3282.09,4349.41)	0.36 (0.36,0.37)	2074360 (1774240,2358850)	296.19 (255.45,336.72)	5405132 (4656883,6153076)
Southeast Asia	7773368 (6600060,8940996)	2885.33 (2480.85,3300.1)	22706789 (19313885,26216287)	3238.01 (2773.46,3719.63)	0.42 (0.41,0.43)	750827 (639280,860351)	247.53 (213.1,282.67)	2050674 (1744848,2371714)
Southern Latin America	1920099 (1659204,2191582)	4128.09 (3574.76,4711.31)	3942908 (3422500,4496778)	4617.51 (4009.6,5266.68)	0.35 (0.32,0.39)	163441 (141110,187399)	348.5 (300.96,399.28)	315715 (274143,360911)
Southern Sub-Saharan Africa	1002298 (855477,1155712)	3626.52 (3109.81,4181.9)	2363886 (2022846,2722791)	3913.26 (3363.99,4509.41)	0.26 (0.25,0.26)	96752 (82606,111033)	315.17 (269.79,363.44)	227947 (195149,260792)
Tropical Latin America	3760894 (3229593,4307856)	4003.22 (3445.06,4588.46)	11648775 (10018943,13328778)	4455.51 (3838.66,5095.51)	0.38 (0.36,0.39)	362189 (310489,413559)	346.45 (297.89,397.37)	1014593 (871428,1161138)
Western Europe	21982300 (19056229,25147170)	3934.19 (3410.31,4494.01)	35172347 (30617674,40161660)	4169.47 (3611.88,4769.26)	0.18 (0.16,0.2)	1708316 (1486092,1964832)	334.48 (289.7,381.98)	2555151 (2238318,2953418)
Western Sub-Saharan Africa	3179864 (2716760,3652589)	3496.16 (2995.24,4022.03)	8096695 (6882245,9292419)	3801.12 (3250.23,4369.9)	0.26 (0.22,0.31)	310356 (264842,356491)	302.36 (259.52,347.66)	831905 (706430,953808)
Location	Incidence (95% uncertainty interval)		DALYs (95% uncertainty interval)
	ASR (2021)	EAPC	Cases (1990)	ASR (1990)	Cases (2021)	ASR (2021)	EAPC
Global (both sex)	353.67 (304.56,402.5)	0.28 (0.26,0.3)	5145339 (2507481,9953258)	127.14 (62.17,246.99)	12019070 (5858108,23267858)	137.59 (67.08,266.87)	0.33 (0.3,0.36)
Global (female)	410.51 (354.2,467.16)	0.29 (0.26,0.32)	3155786 (1543268,6114836)	147.59 (72.23,286.21)	7344351 (3585648,14249004)	160.61 (78.3,311.53)	0.36 (0.33,0.4)
Global (male)	295.15 (253.69,336.09)	0.26(0.25,0.28)	1989553 (966778,3838422)	103.89 (50.66,202.11)	4674719 (2271426,9018855)	112.3 (54.71,217.62)	0.31 (0.28,0.33)
High SDI	386.58 (334.58,439.34)	0.14 (0.1,0.17)	1515496 (742820,2979335)	140.18 (68.63,274.58)	2737095 (1331220,5281509)	149.14 (73.09,292.89)	0.15 (0.12,0.19)
High-middle SDI	357.78 (307.34,407.91)	0.44 (0.39,0.49)	1224964 (596148,2368347)	121.29 (59.1,235.64)	3978378 (1922896,7662185)	138.4 (67.13,267.45)	0.55 (0.49,0.61)
Middle SDI	359.69 (310.37,410.52)	0.33 (0.28,0.38)	1448003 (701826,2778682)	132.02 (64.04,255.23)	2864761 (1411755,5666715)	141.17 (68.46,273.3)	0.38 (0.32,0.45)
Low-middle SDI	321.83 (276.69,366.11)	0.31 (0.3,0.32)	701057 (340832,1342497)	108.81 (53.29,210.14)	614610 (299059,1167797)	120.44 (59.02,231.75)	0.36 (0.35,0.37)
Low SDI	306.06 (264.24,347.69)	0.2 (0.2,0.21)	250983 (122046,481344)	105.33 (51.63,203.7)	1815224 (884403,3477190)	112.74 (55.22,216.52)	0.24 (0.23,0.25)
Andean Latin America	386.9 (333.29,443.77)	0.37 (0.36,0.38)	27640 (13612,53615)	131.2 (64.79,255.6)	88874 (43242,172157)	147.55 (72.02,286.28)	0.42 (0.4,0.43)
Australasia	403.34 (348.15,463.45)	0.34 (0.32,0.37)	31751 (15676,63068)	137.26 (67.72,271.84)	76128 (37728,153473)	154.18 (75.68,309.09)	0.36 (0.33,0.38)
Caribbean	374.83 (324.54,430.56)	0.27 (0.25,0.28)	34448 (16791,66909)	131.6 (64.27,255.98)	77063 (38028,151478)	142.84 (70.48,281.03)	0.3 (0.29,0.32)
Central Asia	239.29 (205.97,274.09)	0.2 (0.18,0.21)	38487 (18666,73649)	81.49 (39.7,157.58)	74350 (36555,143465)	86.71 (42.6,169.71)	0.21 (0.2,0.23)
Central Europe	282.04 (242.21,323.09)	0.23 (0.22,0.23)	143854 (70890,278718)	95.69 (47.26,186)	215177 (106271,422092)	103.21 (50.75,200.27)	0.27 (0.26,0.27)
Central Latin America	383.48 (330.54,435.99)	0.26 (0.25,0.27)	113158 (55122,217637)	131.42 (64.27,254.9)	368356 (179925,711864)	143.61 (70.3,278.58)	0.29 (0.27,0.3)
Central Sub-Saharan Africa	297.77 (256.63,338.31)	0.11 (0.09,0.12)	24860 (12133,47848)	104.1 (50.94,202.82)	66124 (32089,127548)	109.23 (53.33,213.55)	0.14 (0.12,0.16)
East Asia	406.19 (348.6,466.86)	0.46 (0.37,0.56)	1390648 (668816,2682918)	151.13 (72.88,291.29)	3677623 (1773767,7079261)	162.47 (78.29,314.12)	0.5 (0.38,0.61)
Eastern Europe	298.56 (256.44,342.29)	0.25 (0.24,0.27)	281591 (138411,548173)	100.24 (49.43,195.97)	372649 (183304,722199)	108.31 (53.24,209.35)	0.29 (0.28,0.31)
Eastern Sub-Saharan Africa	299.96 (257.48,343.27)	0.23 (0.21,0.24)	80099 (38879,153647)	102.41 (50.13,199.17)	200595 (97952,381074)	110.05 (54.14,212.29)	0.27 (0.25,0.28)
High-income Asia Pacific	458.22 (397.65,522.59)	0.16 (0.12,0.2)	360521 (174917,701559)	174.98 (85.11,341.43)	722000 (353789,1446533)	180.61 (88.02,356.06)	0.16 (0.12,0.2)
High-income North America	389.08 (335.14,444.02)	-0.06 (-0.22,0.11)	482468 (237787,957742)	143.91 (70.69,283.92)	911621 (449630,1807879)	148.92 (73.19,294.55)	-0.15 (-0.34,0.05)
North Africa and Middle East	325.34 (279.63,371.94)	0.35 (0.34,0.35)	191120 (92640,367986)	108.26 (52.86,210.57)	596794 (290214,1148221)	121.33 (59.3,235.41)	0.37 (0.37,0.38)
Oceania	334.57 (286.05,387.71)	0.18 (0.16,0.21)	3889 (1841,7536)	119.61 (57.24,233.97)	10965 (5248,21272)	128.75 (62.39,250.96)	0.2 (0.17,0.23)
South Asia	324.2 (280.14,369.04)	0.31 (0.29,0.32)	669709 (326141,1287959)	108.55 (53.23,210.18)	1872013 (913685,3586458)	120.81 (59.17,232.43)	0.38 (0.37,0.39)
Southeast Asia	274.07 (235.74,314.07)	0.37 (0.36,0.38)	252388 (122330,486666)	92.87 (45.45,179.67)	736837 (356755,1411054)	104.35 (50.89,201.33)	0.43 (0.41,0.44)
Southern Latin America	387.34 (336.02,441.16)	0.31 (0.27,0.34)	61860 (30519,121117)	132.85 (65.49,260.41)	126220 (61981,249769)	148.13 (72.64,291.9)	0.34 (0.31,0.38)
Southern Sub-Saharan Africa	338.1 (290.72,388.21)	0.24 (0.23,0.24)	32212 (15730,62021)	115.9 (56.92,225.01)	75140 (36629,144659)	123.64 (60.67,240.41)	0.23 (0.22,0.23)
Tropical Latin America	380.8 (327.9,434.95)	0.34 (0.32,0.35)	119829 (58969,229561)	126.74 (62.67,245)	369382 (181038,712660)	141.07 (69.25,272.84)	0.38 (0.37,0.39)
Western Europe	357.68 (311.46,406.98)	0.19 (0.17,0.21)	702719 (344993,1392177)	126.28 (61.88,248.06)	1119462 (552620,2232285)	133.93 (65.73,263.33)	0.19 (0.17,0.21)
Western Sub-Saharan Africa	326.72 (280.96,375.13)	0.24 (0.2,0.28)	102089 (49668,195744)	111.44 (54.56,215.6)	261698 (126482,497450)	121.72 (59.6,234.01)	0.28 (0.24,0.33)

Abbreviations: ASR age-standardized rate, DALYs disability-adjusted life years, KOA knee osteoarthritis, EAPC estimated annual percentage change

Fig. 1.

Global distribution of KOA disease burden in 2021. A The ASR of prevalence; B The ASR of incidence; C The ASR of DALYs. Abbreviations: ASR, age-standardized rate; DALYs, disability-adjusted life-years. KOA, knee osteoarthritis

Regional level

In 2021, studies across various regions indicated that the ASPR for KOA was highest in areas with a High SDI, reaching 4,656.7 per 100,000 (95% UI: 4,033.58–5,313.55) (Table 1, Fig. 1A), while low SDI regions reported the lowest at 3,545.62 per 100,000 (95% UI: 3,048.24–4,056.6) (Table 1, Fig. 1A). Temporal trends showed a rising ASPR across all SDI regions, with high-middle SDI areas experiencing the most substantial increase, indicated by an EAPC of 0.53 (95% CI: 0.48 − 0.59). In these regions, the ASPR climbed from 3,772.7 per 100,000 (95% UI: 3,248.87–4,327.23) in 1990 to 4,295.79 per 100,000 (95% UI: 3,685.79–4,926.17) in 2021 (Table 1, Fig. 1A). Conversely, high SDI regions saw the smallest increase, with an EAPC of 0.16 (95% CI: 0.12–0.2) (Table 1, Fig. 1A). Geographically, the highest ASPR was observed in High-income Asia Pacific at 5,573.73 per 100,000 (95% UI: 4,841.29–6,334.33), followed by East Asia at 5,016.78 per 100,000 (95% UI: 4,267.39–5,758.88) (Table 1). In contrast, Central Asia reported the lowest ASPR at 2,700.22 per 100,000 (95% UI: 2,336.33–3,105.34) (Table 1). Overall, all regions except High-income North America showed an upward trend in ASPR over time. East Asia exhibited the most pronounced increase, with an EAPC of 0.49 (95% CI: 0.39 − 0.6), rising from 4,662.94 per 100,000 (95% UI: 3,992.86–5,359.08) in 1990 to 5,016.78 per 100,000 (95% UI: 4,267.39–5,758.88) in 2021 (Table 1). In contrast, High-income North America experienced a decline with an EAPC of −0.11 (95% CI: −0.31–0.08) (Table 1).

The ASIR for KOA was highest in regions with a High SDI, recorded at 386.58 per 100,000 (95% UI: 334.58–439.34), while the lowest was found in Low SDI regions at 306.06 per 100,000 (95% UI: 264.24–347.69) (Table 1, Fig. 1B). High-middle SDI regions showed the most significant increase in ASIR, with an EAPC of 0.44 (95% CI: 0.39 − 0.49) (Table 1, Fig. 1B). Geographically, the regions with the highest ASIR included High-income Asia Pacific at 441.54 per 100,000 (95% UI: 382.22–504.23), East Asia at 406.19 per 100,000 (95% UI: 348.6–466.86), and Australasia at 403.34 per 100,000 (95% UI: 348.15–463.45). Notably, East Asia experienced the most significant upward trend from 1990 to 2021, increasing from 377.47 per 100,000 (95% UI: 324.32–433.77) to 406.19 per 100,000 (95% UI: 348.6–466.86) (Table 1). In contrast, only High-income North America exhibited a downward trend, with an EAPC of −0.06 (95% CI: −0.22–0.11) (Table 1).

The age-standardized DALY rate for KOA was highest in regions with a High SDI at 149.14 per 100,000 (95% UI: 73.09–292.89), and lowest in High SDI regions at 112.74 per 100,000 (95% UI: 55.22–216.52) (Table 1, Fig. 1C). The most significant increase in the age-standardized DALY rate was noted in High-middle SDI regions, with an EAPC of 0.49 (95% CI: 0.49 − 0.61) (Table 1, Fig. 1C). High-income Asia Pacific recorded the highest age-standardized DALY rate at 180.61 per 100,000 (95% UI: 88.02–356.06). Additionally, East Asia demonstrated the most notable upward trend from 1990 to 2021, rising from 151.13 per 100,000 (95% UI: 72.88–291.29) to 162.47 per 100,000 (95% UI: 78.29–314.12) (Table 1). In contrast, only High-income North America showed a decline, with an EAPC of −0.15 (95% CI: −0.34–0.05) (Table 1).

National level

The ASPR of KOA ranges from approximately 2,400 to 6,200 per 100,000 individuals across countries. According to the data, the Republic of Korea exhibits the highest ASPR (6,201.62 per 100,000 persons; 95% UI: 5,389.51–7,093.43), followed by Brunei Darussalam (5,824.76 per 100,000 persons; 95% UI: 5,024.72–6,623.85), Singapore (5,810.72 per 100,000 persons; 95% UI: 5,004.32–6,619.23), Japan (5,331.09 per 100,000 persons; 95% UI: 4,598.71–6,051.31), Taiwan, Province of China (5,263.61 per 100,000 persons; 95% UI: 4,510.48–5,989.03), and China (5,016.52 per 100,000 persons; 95% UI: 4,265.22–5,758.38) (Fig. 1A and Supplementary Table S1). Conversely, Tajikistan (2425.49 per 100,000 persons; 95% UI: 2101.17–2773.29), Kyrgyzstan (2596.13 per 100,000 persons; 95% UI: 2232.44–2957.17), Mongolia (2600.41 per 100,000 persons; 95% UI: 2239.64–2958.12), Uzbekistan (2673.49 per 100,000 persons; 95% UI: 2306.68–3116.56), Armenia (2698.2 per 100,000 persons; 95% UI: 2345.63–3106.37) and Georgia (2722.08 per 100,000 persons; 95% UI: 2322.61–3165.66) exhibited the lowest ASPR (Fig. 1A and Supplementary Table S1). From 1990 to 2021, variations in the change of the ASPR were observed across countries. Notably, Oman (0.68 per 100,000 persons; 95% UI: 0.66–0.7), Equatorial Guinea (0.68 per 100,000 persons; 95% UI: 0.64–0.72), and Thailand (0.59 per 100,000 persons; 95% UI: 0.58–0.61) experienced the most substantial relative increases in ASPR (Supplementary Table S1). In contrast, the only downward trends were observed in the United States of America (−0.12 per 100,000 persons; 95% UI: −0.32–0.08) (Supplementary Table S1).

The ASIR of KOA ranges from approximately 210 to 500 per 100,000 individuals. Notably, Republic of Korea (491.74 per 100,000 persons; 95% UI: 427.19–560.51), Brunei Darussalam (468.07 per 100,000 persons; 95% UI: 407.83–533.21), Singapore (467.22 per 100,000 persons; 95% UI: 405.99–533.81), Japan (441.94 per 100,000 persons; 95% UI: 382.84–503.95), Taiwan (Province of China) (418.79 per 100,000 persons; 95% UI: 361.81–477.52) and Puerto Rico (413.31 per 100,000 persons; 95% UI: 357.6–472.52) have the highest ASIR (Fig. 1B and Supplementary Table S1). Conversely, Tajikistan (218.48 per 100,000 persons; 95% UI: 189.9–250.18), Kyrgyzstan (231.57 per 100,000 persons; 95% UI: 199.58–266.61), Mongolia (231.84 per 100,000 persons; 95% UI: 200.31–266.61), Uzbekistan (237.73 per 100,000 persons; 95% UI: 203.94–274.72), Armenia (239.62 per 100,000 persons; 95% UI: 209.11–276.64) and Georgia (241.18 per 100,000 persons; 95% UI: 206.02–279.95) exhibited the lowest ASIR (Fig. 1B and Supplementary Table S1). The country-specific distribution of ASIR is detailed in Fig. 1B and Supplementary Table S1. Significant changes from 1990 to 2021 were particularly evident in Equatorial Guinea (0.6 per 100,000 persons; 95% UI: 0.56–0.64), Oman (0.6 per 100,000 persons; 95% UI: 0.58–0.62), and Thailand (0.54 per 100,000 persons; 95% UI: 0.52–0.56) (Supplementary Table S1). In contrast, the only downward trends were also observed in the United States of America (−0.07 per 100,000 persons; 95% UI: −0.24–0.11) (Supplementary Table S1).

The age-standardized DALYs rates for the condition ranged from 78 to 200 per 100,000 population. In 2021, Republic of Korea (199.93 per 100,000 persons; 95% UI: 97.42–398.67), Singapore (189.49 per 100,000 persons; 95% UI: 90.34–372.91), Brunei Darussalam (186.75 per 100,000 persons; 95% UI: 91.37–362.74), Japan (172.94 per 100,000 persons; 95% UI: 84.12–338.87), Taiwan (Province of China) (170.98 per 100,000 persons; 95% UI: 81.66–333.52) and China (162.44 per 100,000 persons; 95% UI: 78.35,314.13) had the highest age-standardized DALYs rates (Fig. 1C, Supplementary Table S1). Conversely, Tajikistan (78.21 per 100,000 persons; 95% UI: 38.74–156.27), Mongolia (83.38 per 100,000 persons; 95% UI: 40.98–162.01), Kyrgyzstan (83.75 per 100,000 persons; 95% UI: 40.22–161.24), Uzbekistan (85.96 per 100,000 persons; 95% UI: 42.01–169.34), Armenia (86.74 per 100,000 persons; 95% UI: 42.89–170.72) and Georgia (87.14 per 100,000 persons; 95% UI: 42.82–169.53) exhibited the lowest age-standardized DALYs rates (Fig. 1C, Supplementary Table S1). From 1990 to 2021, the regions with the largest increases in age-standardized DALYs rates were Equatorial Guinea (0.71 per 100,000 persons; 95% UI: 0.67–0.76), Oman (0.68 per 100,000 persons; 95% UI: 0.66–0.7), and Thailand (0.61 per 100,000 persons; 95% UI: 0.59–0.63) (Supplementary Table S1). Conversely, the country with only decrease in age-standardized DALYs rates was United States of America (−0.15 per 100,000 persons; 95% UI: −0.35–0.05). For more information about DALYs, see Fig. 1C and Supplementary Tables S1.

Age and sex patterns

In 2021, no global prevalence, incidence and DALYs rates of KOA was observed before the age of 30 (Fig. 2). The lowest global prevalence, incidence and DALYs rates of KOA were observed in the age of 30 to 34, increasing with age (Fig. 2A). The global prevalence rates increased gradually and reached highest in the age of 80 to 84 in both sex, and then decreased gradually (Fig. 2A). The prevalence rates among females were higher than males at all ages (Fig. 2A). The highest global incidence rate was in the age of 50 to 54 in female, and it was in the age of 60 to 64 in male (Fig. 2B). The incidence rates among females were higher than males in the age before 90, and they were higher among males than females in the age after 90 (Fig. 2B). The DALYs rates were highest in the age of 80 to 84 in both sex, and they were higher in females than those in males at all ages (Fig. 2C).

Fig. 2.

Age- and sex-structured analysis of KOA disease burden in 2021. A The ASR of prevalence; B The ASR of incidence; C The ASR of DALYs. Abbreviations: ASR, age-standardized rate; DALYs, disability-adjusted life-years. KOA, knee osteoarthritis

Overall temporal trends in gender and age structures

From 1990 to 2021, the prevalence number, incidence number and DALYs number in all groups exhibited a general incline. Specifically, the prevalence number, incidence number and rates were consistently higher in females compared to males throughout this period (Fig. 3). An overall increase in prevalence rates, incidence rates and DALYs rates were observed, with an EAPC of 0.3 (95% CI: 0.27 to 0.32), 0.26 (95% CI: 0.25 to 0.28) and 0.31 (95% CI: 0.28 to 0.33) respectively (Fig. 3, Table 1).

Fig. 3.

Overall temporal trends in gender and age structures of KOA disease burden from 1990 to 2021. A The ASR of prevalence; B The ASR of incidence; C The ASR of DALYs. Abbreviations: ASR, age-standardized rate; DALYs, disability-adjusted life-years. KOA, knee osteoarthritis

Temporal joinpoint analysis

Joinpoint regression analysis revealed that global ASPR of KOA was generally on an upward trend from 1990 to 2021 (AAPC = 0.26; 95% CI: 0.23–0.3; P < 0.001) (Fig. 4A, Table 2). The period with the highest increase in ASPR occurred from 2000 to 2008 (APC = 0.55; 95% CI: 0.51–0.59; P < 0.001) (Fig. 4A, Table 2). The ASIR exhibited a global upward trend (AAPC = 0.22; 95% CI: 0.2–0.25; P < 0.001), with the most notable incline during the 2001–2004 period (APC = 0.71; 95% CI: 0.53–0.89; P < 0.001) (Fig. 4B, Table 2). Similarly, the age-standardized DALYs rates followed the same upward trend from 1990 to 2021 (AAPC = 0.26; 95% CI: 0.22–0.29; P < 0.001).

Fig. 4.

Joinpoint regression analysis of the KOA disease burden temporal trends, 1990–2021. A The ASR of prevalence; B The ASR of incidence; C The ASR of DALYs. Abbreviations: ASR, age-standardized rate; DALYs, disability-adjusted life-years. KOA, knee osteoarthritis

Table 2.

Joinpoint regression analysis: trends in KOA age-standardized prevalence, incidence, DALYs rates (per 100,000 persons) in Global and 5 SDI regions, 1990–2021

Location	Prevalence				Incidence				DALYs
	Period	APC (95% CI)	P value	AAPC (95% CI)	Period	APC (95% CI)	P value	AAPC (95% CI)	Period	APC (95% CI)	P value	AAPC (95% CI)
Global	1990–1995	0 (−0.06–0.07)	0.95	0.26 (0.23–0.3)	1990–1997	0.08 (0.05–0.1)	< 0.001	0.22 (0.2–0.25)	1990–1995	0.02 (−0.04–0.08)	0.46	0.26 (0.22–0.29)
	1995–2000	0.22 (0.13–0.31)	< 0.001		1997–2001	0.26 (0.17–0.35)	< 0.001		1995–2000	0.2 (0.11–0.29)	< 0.001
	2000–2008	0.55 (0.51–0.59)	< 0.001		2001–2004	0.71 (0.53–0.89)	< 0.001		2000–2008	0.57 (0.54–0.61)	< 0.001
	2008–2015	0.33 (0.28–0.38)	< 0.001		2004–2015	0.29 (0.28–0.31)	< 0.001		2008–2015	0.33 (0.29–0.38)	< 0.001
	2015–2018	−0.15 (−0.44–0.13)	0.27		2015–2018	−0.08 (−0.26–0.1)	0.36		2015–2018	−0.19 (−0.46–0.09)	0.17
	2018–2021	0.26 (0.12–0.4)	< 0.001		2018–2021	0.09 (0–0.18)	0.05		2018–2021	0.18 (0.04–0.32)	0.01
High SDI	1990–1995	0.09 (−0.05–0.22)	0.18	0.22 (0.16–0.28)	1990–2000	0.28 (0.25–0.3)	< 0.001	0.21 (0.18–0.23)	1990–1995	0.1 (−0.03–0.23)	0.11	0.21 (0.15–0.27)
	1995–2000	0.48 (0.29–0.66)	< 0.001		2000–2005	−0.45 (−0.54–−0.35)	< 0.001		1995–2000	0.45 (0.27–0.63)	< 0.001
	2000–2005	−0.63 (−0.81–−0.44)	< 0.001		2005–2019	0.3 (0.29–0.32)	< 0.001		2000–2005	−0.62 (−0.8–−0.44)	< 0.001
	2005–2010	0.6 (0.41–0.79)	< 0.001		2019–2021	0.84 (0.54–1.14)	< 0.001		2005–2010	0.6 (0.42–0.78)	< 0.001
	2010–2018	0.17 (0.09–0.25)	< 0.001						2010–2018	0.15 (0.08–0.23)	< 0.001
	2018–2021	0.95 (0.65–1.24)	< 0.001						2018–2021	0.88 (0.59–1.17)	< 0.001
High-middle SDI	1990–2000	0.23 (0.21–0.24)	< 0.001	0.42 (0.4–0.45)	1990–2000	0.19 (0.17–0.21)	< 0.001	0.34 (0.32–0.36)	1990–2000	0.23 (0.22–0.24)	< 0.001	0.43 (0.41–0.45)
	2000–2005	1.4 (1.33–1.47)	< 0.001		2000–2005	1.24 (1.14–1.33)	< 0.001		2000–2005	1.43 (1.38–1.48)	< 0.001
	2005–2015	0.42 (0.4–0.44)	< 0.001		2005–2015	0.3 (0.27–0.32)	< 0.001		2005–2015	0.45 (0.44–0.47)	< 0.001
	2015–2019	−0.17 (−0.28–−0.07)	< 0.001		2015–2021	−0.08 (−0.13–−0.04)	< 0.001		2015–2019	−0.22 (−0.3–−0.15)	< 0.001
	2019–2021	0.18 (−0.03–0.39)	0.09						2019–2021	0.12 (−0.03–0.27)	0.11
Middle SDI	1990–1994	−0.36 (−0.43–−0.3)	< 0.001	0.23 (0.2–0.25)	1990–1995	−0.24 (−0.3–−0.18)	< 0.001	0.2 (0.17–0.22)	1990–1998	−0.22 (−0.26–−0.19)	< 0.001	0.23 (0.18–0.27)
	1994–2000	−0.07 (−0.12–−0.03)	< 0.001		1995–2000	0.02 (−0.06–0.11)	0.61		1998–2001	0.34 (−0.01–0.68)	0.06
	2000–2005	1.31 (1.24–1.38)	< 0.001		2000–2005	1.05 (0.97–1.14)	< 0.001		2001–2005	1.45 (1.27–1.63)	< 0.001
	2005–2015	0.39 (0.37–0.41)	< 0.001		2005–2015	0.31 (0.28–0.33)	< 0.001		2005–2015	0.39 (0.35–0.42)	< 0.001
	2015–2019	−0.42 (−0.53–−0.32)	< 0.001		2015–2021	−0.19 (−0.23–−0.14)	< 0.001		2015–2019	−0.48 (−0.65–−0.31)	< 0.001
	2019–2021	0.13 (−0.08–0.34)	0.21				< 0.001		2019–2021	0.04 (−0.31–0.39)	0.82
Low-middle SDI	1990–1996	0.24 (0.22–0.26)	< 0.001	0.33 (0.31–0.35)	1990–1996	0.19 (0.19–0.2)	< 0.001	0.29 (0.28–0.3)	1990–1996	0.25 (0.23–0.27)	< 0.001	0.33 (0.31–0.35)
	1996–2001	0.36 (0.32–0.4)	< 0.001		1996–2001	0.32 (0.3–0.34)	< 0.001		1996–2001	0.37 (0.32–0.41)	< 0.001
	2001–2004	0.48 (0.34–0.61)	< 0.001		2001–2004	0.44 (0.39–0.5)	< 0.001		2001–2004	0.49 (0.35–0.63)	< 0.001
	2004–2019	0.34 (0.34–0.35)	< 0.001		2004–2010	0.26 (0.25–0.27)	< 0.001		2004–2019	0.36 (0.35–0.36)	< 0.001
	2019–2021	0.2 (0.07–0.33)	< 0.001		2010–2019	0.34 (0.34–0.35)	< 0.001		2019–2021	0.06 (−0.08–0.19)	0.39
					2019–2021	0.16 (0.11–0.21)	< 0.001
Low SDI	1990–1996	0.11 (0.09–0.14)	< 0.001	0.21 (0.2–0.23)	1990–1996		< 0.001	0.2 (0.18–0.21)	1990–2000	0.15 (0.15–0.16)	< 0.001	0.22 (0.22–0.23)
	1996–2001	0.21 (0.16–0.25)	< 0.001		1996–2001	0.1 (0.08–0.13)	< 0.001		2000–2004	0.38 (0.33–0.42)	< 0.001
	2001–2004	0.37 (0.23–0.51)	< 0.001		2001–2004	0.21 (0.16–0.25)	< 0.001		2004–2019	0.24 (0.24–0.25)	< 0.001
	2004–2016	0.2 (0.19–0.21)	< 0.001		2004–2013	0.32 (0.17–0.46)	< 0.001		2019–2021	0.14 (0.05–0.23)	< 0.001
	2016–2021	0.26 (0.23–0.29)	< 0.001		2013–2021	0.16 (0.15–0.18)	< 0.001

Abbreviations: SDI sociodemographic index, DALYs disability-adjusted life years, KOA knee osteoarthritis, APC annual percent change, AAPC average annual percent change

An analysis of the 5 SDI regions showed that the ASPR, ASIR and the age-standardized DALYs rates of 5 SDI regions also generally on an upward trend from 1990 to 2021 (Fig. 4, Table 2). The ASPR (AAPC = 0.42; 95% CI: 0.4–0.45; P < 0.001), ASIR (AAPC = 0.34; 95% CI: 0.32–0.36; P < 0.001) and the age-standardized DALYs rates (AAPC = 0.43; 95% CI: 0.41–0.45; P < 0.001) of High-middle SDI region showed the most notable incline. Noteworthy, the ASPR (APC = −0.63; 95% CI: −0.81–−0.44; P < 0.001), ASIR (APC = −0.45; 95% CI: −0.54–−0.35; P < 0.001) and the age-standardized DALYs rates (APC = −0.62; 95% CI: −0.8–−0.44; P < 0.001) of high SDI region showed a decline from 2000 to 2005.

Age-period-cohort analysis

The results of age-period-cohort analysis on ASR (age-standardized rate) of KOA prevalence, incidence and DALYs were illustrated in Fig. 5. After controlling for the period and birth cohort effects, the age effect demonstrated a remarkable impact on the risk of KOA prevalence, incidence and DALYs. The relative prevalence, incidence and DALYs risks showed the trends of rising first and then falling, with the highest risk in those aged 70–75 years, 55–60 years and 70–75 years, respectively (Fig. 5, Table 3).

Fig. 5.

The effects of age, period, and birth cohort on the relative risk of KOA. A The ASR of prevalence; B The ASR of incidence; C The ASR of DALYs. Abbreviations: ASR, age-standardized rate; DALYs, disability-adjusted life-years. KOA, knee osteoarthritis

Table 3.

RRs of KOA prevalence, incidence and DALYs for both sexes due to age, period, and birth cohort effects

Factor	Prevalence		Incidence		DALYs
	RR (95% CI)	P	RR (95% CI)	P	RR (95% CI)	P
Age (years)
30–34	0.023 (0.023–0.023)	< 0.001	0.151 (0.15–0.151)	< 0.001	0.024 (0.024–0.025)	< 0.001
35–39	0.202 (0.202–0.202)	< 0.001	0.593 (0.592–0.595)	< 0.001	0.21 (0.21–0.211)	< 0.001
40–44	0.536 (0.536–0.536)	< 0.001	1.096 (1.094–1.098)	< 0.001	0.557 (0.556–0.559)	< 0.001
45–49	0.989 (0.989–0.99)	< 0.001	1.625 (1.623–1.628)	< 0.001	1.025 (1.024–1.027)	< 0.001
50–54	1.46 (1.459–1.46)	< 0.001	1.869 (1.866–1.871)	< 0.001	1.508 (1.505–1.51)	< 0.001
55–59	1.805 (1.805–1.806)	< 0.001	1.895 (1.893–1.897)	< 0.001	1.854 (1.852–1.856)	< 0.001
60–64	2.02 (2.02–2.021)	< 0.001	1.855 (1.854–1.857)	< 0.001	2.059 (2.057–2.061)	< 0.001
65–69	2.131 (2.13–2.131)	< 0.001	1.775 (1.773–1.777)	< 0.001	2.15 (2.147–2.153)	< 0.001
70–74	2.172 (2.171–2.172)	< 0.001	1.633 (1.631–1.636)	< 0.001	2.165 (2.163–2.168)	< 0.001
75–79	2.131 (2.13–2.131)	< 0.001	1.429 (1.427–1.432)	< 0.001	2.096 (2.093–2.1)	< 0.001
80–84	2.02 (2.019–2.02)	< 0.001	1.141 (1.139–1.144)	< 0.001	1.958 (1.954–1.962)	< 0.001
85–89	1.824 (1.823–1.824)	< 0.001	0.782 (0.78–0.785)	< 0.001	1.736 (1.732–1.741)	< 0.001
90–94	1.569 (1.568–1.57)	< 0.001	0.554 (0.551–0.556)	< 0.001	1.465 (1.46–1.47)	< 0.001
95–99	1.327 (1.325–1.328)	< 0.001	0.467 (0.463–0.47)	< 0.001	1.214 (1.207–1.221)	< 0.001
Period
1992–1996	0.767 (0.766–0.767)	< 0.001	0.926 (0.925–0.928)	< 0.001	0.773 (0.772–0.774)	< 0.001
1997–2001	0.848 (0.848–0.848)	< 0.001	0.952 (0.951–0.952)	< 0.001	0.852 (0.851–0.853)	< 0.001
2002–2006	0.951 (0.951–0.951)	< 0.001	0.991 (0.99–0.991)	< 0.001	0.953 (0.952–0.953)	< 0.001
2007–2011	1.065 (1.065–1.065)	< 0.001	1.023 (1.023–1.024)	< 0.001	1.064 (1.063–1.065)	< 0.001
2012–2016	1.18 (1.18–1.18)	< 0.001	1.05 (1.049–1.051)	< 0.001	1.174 (1.173–1.175)	< 0.001
2017–2021	1.288 (1.288–1.288)	< 0.001	1.066 (1.065–1.067)	< 0.001	1.275 (1.274–1.276)	< 0.001
Birth cohort
1897–1901	2.209 (2.202–2.217)	< 0.001	1.148 (1.116–1.18)	< 0.001	2.144 (2.1–2.189)	< 0.001
1902–1906	2.05 (2.047–2.053)	< 0.001	1.131 (1.118–1.145)	< 0.001	1.995 (1.977–2.014)	< 0.001
1907–1911	1.893 (1.891–1.895)	< 0.001	1.118 (1.11–1.126)	< 0.001	1.848 (1.837–1.858)	< 0.001
1912–1916	1.746 (1.744–1.747)	< 0.001	1.106 (1.1–1.112)	< 0.001	1.71 (1.702–1.717)	< 0.001
1917–1921	1.594 (1.593–1.595)	< 0.001	1.101 (1.096–1.107)	< 0.001	1.565 (1.56–1.571)	< 0.001
1922–1926	1.447 (1.446–1.448)	< 0.001	1.092 (1.087–1.096)	< 0.001	1.426 (1.422–1.431)	< 0.001
1927–1931	1.303 (1.302–1.303)	< 0.001	1.069 (1.065–1.073)	< 0.001	1.288 (1.284–1.292)	< 0.001
1932–1936	1.18 (1.179–1.18)	< 0.001	1.038 (1.035–1.042)	< 0.001	1.17 (1.167–1.173)	< 0.001
1937–1941	1.073 (1.072–1.073)	< 0.001	1.009 (1.007–1.012)	< 0.001	1.068 (1.065–1.07)	< 0.001
1942–1946	0.984 (0.983–0.984)	< 0.001	0.987 (0.985–0.99)	< 0.001	0.982 (0.981–0.984)	< 0.001
1947–1951	0.903 (0.903–0.904)	< 0.001	0.968 (0.967–0.97)	< 0.001	0.906 (0.904–0.907)	< 0.001
1952–1956	0.831 (0.83–0.831)	< 0.001	0.949 (0.948–0.95)	< 0.001	0.836 (0.835–0.837)	< 0.001
1957–1961	0.757 (0.757–0.757)	< 0.001	0.929 (0.928–0.93)	< 0.001	0.765 (0.764–0.766)	< 0.001
1962–1966	0.704 (0.704–0.704)	< 0.001	0.926 (0.926–0.927)	< 0.001	0.714 (0.713–0.715)	< 0.001
1967–1971	0.664 (0.663–0.664)	< 0.001	0.946 (0.945–0.946)	< 0.001	0.676 (0.675–0.677)	< 0.001
1972–1976	0.602 (0.602–0.602)	< 0.001	0.938 (0.937–0.939)	< 0.001	0.615 (0.614–0.617)	< 0.001
1977–1981	0.534 (0.534–0.534)	< 0.001	0.902 (0.901–0.904)	< 0.001	0.548 (0.546–0.549)	< 0.001
1982–1986	0.479 (0.478–0.479)	< 0.001	0.869 (0.867–0.871)	< 0.001	0.493 (0.491–0.495)	< 0.001
1987–1991	0.437 (0.436–0.439)	< 0.001	0.852 (0.849–0.855)	< 0.001	0.452 (0.446–0.459)	< 0.001

Abbreviations: DALYs disability-adjusted life years, KOA knee osteoarthritis, RR relative risk, CI confidence interval

After controlling for the age and period effects, birth cohort effect imposed a significant impact on the risk of KOA prevalence, incidence and DALYs. The birth cohort effect presented a higher risk of prevalence, incidence and DALYs in earlier birth cohorts compared with later cohorts, with the relative risk (RR) continuously decreasing from the 1897–1901 cohort to the 1987–1991 cohort (Fig. 5, Table 3).

After controlling for the age and birth cohort effects, significant impacts of period effect on the risk of KOA prevalence, incidence and DALYs were observed. The period effect on prevalence, incidence and DALYs risk presented both increasing trends, with 1.68, 1.15 and 1.65 times increment of RR from period 1992 to period 2017, respectively. The highest risk of incidence and prevalence were observed in period 2017 (Fig. 5, Table 3).

Decomposition analysis

The population was the primary contributor to global and five SDI regions (Fig. 6, Supplementary Table S2). Aging, population growth, and epidemiological changes contributed 15.6%, 74.72%, and 9.69% to the global increase in ASPR, respectively (Fig. 6A, Supplementary Table S2). Notably, the highest contributions for aging, population growth, and epidemiological changes were observed in the high SDI quintile (28.53%), low SDI quintile (97.65%), and high-middle SDI quintile (16.45%) (Supplementary Table S2). Similarly, these factors accounted for 9.16%, 81.82%, and 9.02% of the global increase in ASIR, respectively (Fig. 6B, Supplementary Table S2), with the most significant contributions in the high SDI quintile (16.09%), low SDI quintile (95.78%), and high-middle SDI quintile (14.98%) (Supplementary Table S2). Lastly, aging, population growth, and epidemiological changes represented 15.29%, 75.07%, and 9.64% of the global increase in age-standardized DALY rates (Fig. 6C, Supplementary Table S2), with the most pronounced contributions occurring in the high SDI quintile (28.11%), low SDI quintile (97.28%), and high-middle SDI quintile (16.76%) (Supplementary Table S2).

Fig. 6.

Decomposition Analysis of KOA disease burden in 2021. A The ASR of prevalence; B The ASR of incidence; C The ASR of DALYs. Abbreviations: ASR, age-standardized rate; DALYs, disability-adjusted life-years. KOA, knee osteoarthritis

Cross-country inequality analysis

An analysis of age-standardised DALY rates was conducted on 204 countries. As demonstrated by the slope index of inequality, there was an excess of 18.33 (95% CI: 11.26–25.4) DALYs per 100,000 between countries with the highest and lowest SDI in 1990, and it decreased to 16.81 (95% CI: 9.01–24.62) in 2021, which meaned that the absolute SDI-associated inequalities in KOA burden was declining (Fig. 7A). The concentration curve is below the diagonal line, indicating that the burden is concentrated in countries with high socioeconomic development levels (Fig. 7B). However, the concentration index, a measure of relative gradient inequality, presented 0.04 (95% CI: 0.03–0.06) in 1990 and 0.06 (95% CI: 0.05–0.07) in 2021, indicating a slight increased relative SDI-associated inequalities (Fig. 7B).

Fig. 7.

Cross-country inequality analysis of KOA in 1990 and 2021. A SDI-related health inequality regression; B concentration curves for the DALYs of KOA worldwide. Abbreviations: SDI, sociodemographic index; DALYs, disability-adjusted life-years; KOA, knee osteoarthritis

Predictive analysis of KOA burden to 2045

Predictive analysis results for KOA indicated that by 2045, case numbers for prevalence, incidence, and DALYs were projected to rise to 658,088,384.48 (322,110,040.98–994,066,727.98), 47,256,502.97 (23,440,017.7–71,072,988.23) and 20,517,479.78 (10,056,930.68–30,978,028.88), respectively (Fig. 8, Table S3). Meanwhile, the ASR for prevalence, incidence, and DALYs were also expected to incline to 4630.66 (2261.44–6999.87), 373.92 (184.17–563.66) and 145.4 (71.08–219.73), respectively (Fig. 8, Table S3).

Fig. 8.

Predictive Analysis of KOA Burden to 2045. A The predicted case number and ASR of prevalence to 2045; B The predicted case number and ASR of incidence to 2045; C The predicted case number and ASR of DALYs to 2045. Abbreviations: ASR, age-standardized rate; DALYs, disability-adjusted life-years. KOA, knee osteoarthritis

Discussion

This study presents the latest data on the global, regional, and national prevalence, incidence, and DALYs associated with KOA from 1990 to 2021. It offers a comprehensive analysis through trend evaluation, decomposition, inequality assessment, and predictive modeling. Although the prevalence, incidence and DALYs of KOA vary across countries, the overall global burden has risen over this period, with females experiencing a greater impact than males. The most significant increase occurred between 2000 and 2008. Decomposition analysis highlighted population growth and aging as main factors behind the rising KOA burden. Inequality analysis showed that high-SDI countries disproportionately carried this burden, though these disparities have diminished over time. Moreover, both the total number of cases and the rates are projected to continue rising, presenting an ongoing challenge for the effective management of KOA in the coming decades.

Compared with findings from GBD 2019, our GBD 2021-based study reveals both sustained and emerging patterns. While the 2019 data highlighted an overall upward trend, the 2021 data show a more nuanced picture. Notably, we found that the ASRs of KOA in East Asian countries such as South Korea and China have continued to rise sharply, contrasting with stable or even declining ASRs in many High SDI countries. For instance, the United States showed a downward trend in ASR for both prevalence and DALYs—likely reflecting successful public health campaigns and improved clinical management. These contrasting trajectories between East Asia and High SDI regions emphasize the growing heterogeneity in KOA burden, highlighting the novel contribution of our study in identifying shifting regional dynamics beyond what was reported in GBD 2019.

With increasing human life expectancy, a rising obese population, higher rates of joint trauma, advancements in imaging technologies, and continuous development in biomarker and gene detection, the diagnosis of OA has significantly improved. However, the personal and societal economic burden caused by KOA has also increased dramatically [3]. In 2021, KOA accounted for 374.7 million prevalent cases, 30.8 million incident cases, and 12 million DALYs globally, with the highest burden in High-income Asia Pacific and the lowest in Central Asia. Notably, countries such as the Republic of Korea, Japan, Singapore, and Mainland China showed higher ASRs, while parts of Central Asia exhibited relatively lower burdens. These variations are likely influenced by multiple factors.

Firstly, population aging is a critical driver for the increasing KOA burden. The prevalence of KOA rises significantly with age, so countries with a higher proportion of elderly populations naturally bear a heavier burden. The Republic of Korea, for instance, is among the fastest aging countries globally, with studies reporting KOA prevalence exceeding 30% among women aged over 65 years [27–29]. Secondly, healthcare resources and the robustness of health systems affect disease diagnosis and reporting levels. Developed countries with better healthcare access and early screening programs tend to report higher detection rates. For example, the universal health coverage systems in Japan and Singapore facilitate early diagnosis and disease management [30]. Furthermore, lifestyle changes such as urbanization and rising obesity rates exacerbate KOA burden. Obesity is a major risk factor for KOA, and the global increase in obesity directly impacts KOA incidence trends [31]. The rise in ASRs observed in rapidly urbanizing countries such as Thailand and Oman reflects this epidemiological transition [32, 33]. Conversely, the lower burden in some Central Asian countries may be attributed to limited healthcare resources, under-recognition of the disease, and imperfect statistical systems leading to underestimation of the actual burden. Rural healthcare shortages in these regions often mean patients seek medical attention only after severe functional impairment, impacting data accuracy [34]. It is also noteworthy that the KOA burden in the United States is showing a decreasing trend, possibly due to widespread health interventions [3]. In summary, the differences in KOA disease burden among countries result from a complex interplay of demographic structure, healthcare systems, lifestyle factors, and socioeconomic conditions. Tailored prevention and control strategies should be developed according to the specific circumstances of each region.

To better understand the relationship between age and KOA, the population was divided into 5-year age groups to account for the impact of aging. The analysis revealed no cases of KOA in individuals younger than 30, consistent with previous studies [35]. Female sex is a well-established clinical risk factor for KOA [36], and numerous studies have shown a higher prevalence of KOA among females compared to males [37–39]. Our findings also indicated that females had higher prevalence rates at all ages. The highest global incidence was observed in women aged 50 to 54 and in men aged 60 to 64. Silverwood et al. [40] highlighted the higher obesity prevalence in women (18% vs. 10% in men), which may contribute to the greater KOA burden in females. Additionally, factors such as the decline in estrogen levels around menopause [34, 41], women’s occupational roles [3], and the generally stronger joint support in men [42] could explain the earlier peak incidence of KOA in women. Targeted early screening and health education programs could be beneficial for specific age groups in high-burden regions [3]. Older age is another recognized clinical risk factor for KOA [43]. Our results showed that the highest prevalence and DALY rates occurred in individuals aged 80 to 84 in both sexes. Overall, these findings align with those of previous studies.

When the overall trend was divided into subsegments, the most significant increase in the ASR of prevalence and DALYs of KOA occurred between 2000 and 2008, while the most notable rise in the ASR of incidence was observed from 2001 to 2004. Similarly, a study using GBD 2019 data showed that the steepest increase in the ASR of prevalence, incidence, and DALYs of OA occurred between 2000 and 2005, likely due to global population aging and the obesity epidemic [11]. The growing aging population and increasing obesity rates, combined with inadequate management in the early 2000 s, may have contributed to the rising KOA burden during this time [44].

However, the greatest decline in the ASR of prevalence and DALYs of KOA between 2000 and 2005 was seen in High SDI regions. This decrease may be linked to the development of effective prevention, management, and treatment programs for KOA in these areas. In 2002, the United Nations convened the Second World Assembly on Ageing in Madrid, adopting the Madrid International Plan of Action on Ageing, which promoted changes in policies and practices to support healthy aging. Additionally, the World Health Organization called for obesity prevention and the adoption of healthy lifestyles in 2005 [11]. The successful implementation of these global initiatives, particularly in High SDI countries, likely contributed to the subsequent reduction in the KOA burden. In contrast, the sharpest increases in the ASR of prevalence, incidence, and DALYs of KOA from 2000 to 2005 were seen in High-middle and Middle SDI regions, and from 2001 to 2004 in Low-middle and Low SDI regions. This suggests that countries in these regions did not prioritize the implementation of these policies, highlighting the need for them to adopt lessons from the High SDI countries’ successful approaches.

Examining the effects of age, period, and birth cohort on KOA can enhance our understanding of the disease’s epidemiology. Ageing is widely recognized as a major risk factor for KOA, contributing to its increasing burden. The relative incidence risk rises with age, peaking in the 55–59 years age group before declining in older groups. Previous studies using GBD data similarly found the highest incidence of OA in this age group [11, 45], suggesting that individuals in this range should receive targeted prevention, management, and treatment. Additionally, the relative prevalence and DALY risks followed a pattern of rising and then falling, with the highest rates seen in the 70–74 years age group, consistent with earlier research on OA [11].

Regarding period effects, the relative risks for prevalence, incidence, and DALYs increased over time, likely due to rising risk factors for KOA, such as obesity, increased physical activity, and improved health registration systems [46, 47]. As for birth cohort effects, decreasing trends in relative risks for prevalence, incidence, and DALYs were observed with each successive cohort, indicating that individuals born earlier have a higher risk for KOA than those born later. This may be due to differences in the social, economic, and cultural environments experienced during early life, with later-born individuals benefiting from better health education and medical care. Decomposition analysis of KOA prevalence, incidence, and DALYs revealed that the overall burden increase was primarily driven by demographic changes, particularly population growth and aging, across all levels of development. In regions with higher SDI, population aging had a greater impact, while population growth had a smaller effect.

The cross-country inequality analysis revealed that a disproportionate KOA burden was concentrated in high SDI countries, consistent with previous studies indicating that nations with higher sociodemographic development bear a greater share of the OA burden [11, 45]. This inequality may be largely due to aging populations and the obesity epidemic in these regions. Additionally, underdiagnosis of KOA in low SDI countries, caused by poor healthcare infrastructure, could also contribute to this disparity. As a result, high SDI countries should focus on developing more effective prevention and treatment strategies for KOA, particularly addressing aging and obesity. Conversely, lower SDI countries face significant challenges due to rapid population growth and a lack of medical resources. Given the observed increases in KOA burden in Low and Middle SDI regions, especially in countries like Equatorial Guinea where the ASR surge was particularly sharp, targeted interventions are essential. Policymakers in these regions should prioritize the development and implementation of affordable screening programs, community-based exercise initiatives, and anti-obesity campaigns. In addition, increasing investments in primary care infrastructure and musculoskeletal training for healthcare professionals can facilitate early diagnosis and management. International support—particularly through World Health Organization, United Nations Development Program, and regional collaborations—should focus on technical assistance, health system strengthening, and equitable access to assistive technologies such as knee braces and joint replacement surgery. Notably, the absolute inequalities in KOA burden associated with SDI have been decreasing, suggesting that current measures for KOA prevention, management, and treatment are proving effective. However, a slight rise in relative SDI-associated inequalities was observed, indicating that further research is needed to explore the underlying causes.

Notably, the total number of cases and ASR for all three metrics is expected to rise, signaling a significant disease burden and emerging challenges in KOA control and management. A study based on GBD 2019 also predicted an increase in prevalence, incidence, and DALYs of OA by 2035, with population growth and aging identified as key factors contributing to this rise [11]. With the global population continuing to grow and age more severely by 2045, it is crucial for all countries to strengthen their healthcare systems to adequately address these demographic changes.

This study has several limitations. First, the GBD database relies primarily on a combination of published studies, health surveys, and modeled estimates rather than uniform primary data collection across countries. This introduces heterogeneity in diagnostic definitions, case ascertainment methods, and reporting quality. Second, underreporting remains a critical issue in low-resource settings. In many low-SDI countries, KOA may not be diagnosed until severe functional impairment occurs, and healthcare access is often limited. As a result, DALY estimates in these regions may be underestimated. Third, our study did not fully account for the potential influence of the COVID-19 pandemic on KOA data from 2020–2021. The pandemic disrupted routine healthcare services, reduced elective surgeries including joint replacements, and limited access to physiotherapy and outpatient care. These disruptions could have led to both underdiagnosis and increased disability among KOA patients, particularly in vulnerable groups [48]. Therefore, pandemic-related factors may confound the 2020–2021 data, and further studies are needed to isolate and understand these effects. Lastly, while the GBD methodology includes advanced statistical modeling and correction factors, the accuracy of its outputs still depends on the quality and completeness of input data. Regional gaps in data—especially from low-income countries—introduce uncertainty into our estimates. Sensitivity analyses and further validation using prospective cohort data and clinical registries are warranted to improve the robustness of future KOA burden assessments.

In conclusion, the global burden of KOA increased steadily from 1990 to 2021, with a greater burden on females compared to males. The most rapid growth occurred between 2000 and 2008, driven primarily by population growth and aging, which are expected to further escalate the number of KOA cases until 2045. Countries with high SDI have borne a disproportionately high burden, and SDI-related inequalities have worsened over time. Current KOA management and treatment face significant challenges. Strengthening healthcare systems, promoting healthy lifestyles, and addressing socioeconomic disparities are essential steps in reducing the global burden of KOA. Global health policymakers should prioritize targeted interventions and adaptable strategies to enhance personalized healthcare systems that meet the specific needs of each country.

Abbreviations

OA

Osteoarthritis

KOA

Knee osteoarthritis

GBD

Global Burden of Disease

UI

Uncertainty interval

DALY

Disability-adjusted life year

SDI

Sociodemographic index

APC

Annual percent change

AAPC

Average annual percent change

ASPR

Age-standardized prevalence rate

EAPC

Estimated annual percentage change

ASIR

Age-standardized incidence rate

RR

Relative risk

ASR

Age-standardized rate

Authors’ contributions

OY contributed to conceptualization, data curation, formal analysis, investigation, methodology, software, resources, validation, visualization and writing–review & editing. DM contributed to funding acquisition, project administration, supervision, writing–original draft and writing–review & editing.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The data resources from the GBD study 2021 can be accessed online through the Global Health Data Exchange (GHDx) query tool at http://ghdx.healthdata.org/gbd-results-tool.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.